Available via license: CC BY 4.0
Content may be subject to copyright.
Production of ammonia makes Venusian clouds
habitable and explains observed cloud-level
chemical anomalies
William Bains
a,b
, Janusz J. Petkowski
a
, Paul B. Rimmer
c,d,e
, and Sara Seager
a,f,g,1
a
Department of Earth, Atmospheric, and Planetary Sciences, Massachusetts Institute of Technology,Cambridge, MA 02139;
b
School of Physics & Astronomy,
Cardiff University,Cardiff CF24 3AA, United Kingdom;
c
Department of Earth Sciences, University of Cambridge, Cambridge CB2 3EQ, United Kingdom;
d
Cavendish Laboratory, University of Cambridge, Cambridge CB3 0HE, United Kingdom;
e
Medical Research Council, Laboratory of Molecular Biology,
Cambridge CB2 0QH, United Kingdom;
f
Department of Physics, Massachusetts Institute of Technology, Cambridge, MA 02139; and
g
Department of
Aeronautics and Astronautics, Massachusetts Institute of Technology, Cambridge, MA 02139
Contributed by Sara Seager; received November 6, 2021; accepted November 11, 2021; reviewed by David Grinspoon and Thomas Puzia
The atmosphere of Venus remains mysterious, with many out-
standing chemical connundra. These include the unexpected pres-
ence of ∼10 ppm O
2
in the cloud layers, an unknown composition
of large particles in the lower cloud layers, and hard to explain
measured vertical abundance profiles of SO
2
and H
2
O. We propose
a hypothesis for the chemistry in the clouds that largely addresses
all of the above anomalies. We include ammonia (NH
3
), a key com-
ponent that has been tentatively detected both by the Venera
8 and Pioneer Venus probes. NH
3
dissolves in some of the sulfuric
acid cloud droplets, effectively neutralizing the acid and trapping
dissolved SO
2
as ammonium sulfite salts. This trapping of SO
2
in
the clouds, together with the release of SO
2
below the clouds as
the droplets settle out to higher temperatures, explains the verti-
cal SO
2
abundance anomaly. A consequence of the presence of
NH
3
is that some Venus cloud droplets must be semisolid ammo-
nium salt slurries, with a pH of ∼1, which matches Earth acidophile
environments, rather than concentrated sulfuric acid. The source
of NH
3
is unknown but could involve biological production; if so,
then the most energy-efficient NH
3
-producing reaction also creates
O
2,
explaining the detection of O
2
in the cloud layers. Our model
therefore predicts that the clouds are more habitable than previ-
ously thought, and may be inhabited. Unlike prior atmospheric
models, ours does not require forced chemical constraints to match
the data. Our hypothesis, guided by existing observations, can be
tested by new Venus in situ measurements.
Venus jclouds jatmospheric chemistry jastrobiology jhabitability
Venus is often called Earth’s sister planet because of its simi-
lar mass and size to Earth. Yet, owing, in part, to the
greenhouse effect from its massive CO
2
atmosphere, Venus’s
surface temperature is higher than 700 K—too hot for life of
any kind. The Venusian surface is therefore a complete contrast
to Earth’s temperate surface and rich surface biosphere. None-
theless, scientists have been speculating on Venus as a habitable
world for over half a century (1–7). Such speculations are based
on the Earth-like temperature and pressure at the altitudes of
48 km to 60 km above the surface (8, 9).
Venus is perpetually shrouded in an ∼20-km-deep layer of
clouds, including the temperate atmosphere layers at 48 km to
60 km. The prevailing consensus is that the clouds of Venus are
made from droplets of concentrated sulfuric acid. This conclu-
sion is inferred from the presence of small amounts of sulfuric
acid vapor in the atmosphere (10, 11) and the refractive index
of cloud droplets (12, 13). While the clouds are often described
as “temperate” or “clement,” such a statement is misleading
when it comes to habitability. If the cloud particles are actually
made of concentrated sulfuric acid, then it is difficult to imag-
ine how life chemically similar to life on Earth could survive
(7, 14). Specifically, the aggressive chemical properties of sulfu-
ric acid and the extremely low atmospheric water content
(14, 15) are orders of magnitude more acidic and 50 to 100
times drier than any inhabited extreme environment on Earth.
Overview of Venusian Atmosphere Anomalies
Despite over 50 y of remote and local observation, Venus’s
atmosphere has a number of lingering anomalies with either
poor model fits or no explanations (16, 17).
One such long-standing mysterious feature of the atmo-
sphere, which is not well explained by current atmospheric
chemistry models, is the abundance profile of water vapor and
SO
2
in and above the cloud layers (17–19).
Observations show that H
2
O persists throughout the atmo-
sphere, while the SO
2
is observed in parts per million abundan-
ces below the clouds and parts per billion abundance above the
clouds. Yet, expectations are very different. The primary source
of SO
2
and H
2
O in the atmosphere of Venus is volcanism. As
the gases are released from volcanoes, they are uniformly
mixed vertically throughout the atmosphere. At very high alti-
tudes in the atmosphere, around 70 km, SO
2
and H
2
O are effi-
ciently destroyed by ultraviolet (UV) radiation. However, the
observed SO
2
and H
2
O abundance profiles deviate from the
uniform distribution, notably, such that SO
2
shows significant
depletion in the cloud layers and H
2
O is present above the
cloud layers.
Significance
This research provides a transformative hypothesis for the
chemistry of the atmospheric cloud layers of Venus while
reconciling decades-long atmosphere anomalies. Our model
predicts that the clouds are not entirely made of sulfuric
acid, but are partially composed of ammonium salt slurries,
which may be the result of biological production of ammo-
nia in cloud droplets. As a result, the clouds are no more
acidic than some extreme terrestrial environments that har-
bor life. Life could be making its own environment on
Venus. The model’s predictions for the abundance of gases
in Venus’atmosphere match observation better than any
previous model, and are readily testable.
Author contributions: W.B. and P.B.R. designed research; W.B. and P.B.R. performed
research; W.B., J.J.P., P.B.R., and S.S. analyzed data; and W.B., J.J.P., P.B.R., and S.S.
wrote the paper.
Reviewers: D.G., Planetary Science Institute; and T.P., Institute of Astrophysics,
Pontificia Universidad Catolica.
The authors declare no competing interest.
This open access article is distributed under Creative Commons Attribution License 4.0
(CC BY).
1
To whom correspondence may be addressed. Email: seager@mit.edu.
This article contains supporting information online at http://www.pnas.org/lookup/
suppl/doi:10.1073/pnas.2110889118/-/DCSupplemental.
Published December 20, 2021.
PNAS 2021 Vol. 118 No. 52 e2110889118 https://doi.org/10.1073/pnas.2110889118 j
1of10
EARTH, ATMOSPHERIC,
AND PLANETARY SCIENCES
Downloaded by guest on December 20, 2021
Previous consensus models explained the SO
2
profile by sug-
gesting that SO
2
is photochemically oxidized to SO
3
, which
then reacts with water to form sulfuric acid in the clouds:
CO2þhν!CO þO
SO2þOþM!SO3þM
SO3þ2H2O!H2SO4þH2O
SO2þH2OþCO2!H2SO4þCO:
However, as there is 5×more SO
2
than H
2
O, this chemistry
should strip all the water out of the cloud layer, and addition-
ally react with and prevent water from reaching and accumulat-
ing above the clouds as well, while only reducing SO
2
by 20%,
not the 99.9% observed (20). Previous models provide a
numerical fix to match the observations, arbitrarily removing
SO
2
or artificially keeping the water abundance constant
(21, 22).
Another mystery is the presence of O
2
in the clouds (23, 24),
as there is no known process for O
2
formation in the cloud
layers (discussed further below). Finally, the SO
2
,O
2
, and H
2
O
anomalies, together with other trace atmospheric gas abundan-
ces, form a chemical disequilibrium in the clouds of Venus
(25–27).
A more tentative but intriguing anomaly is that of the detec-
tion of NH
3
in and below the cloud layers. NH
3
was tentatively
detected both by the Venera 8 chemical probe (28) and in rean-
alyzed Pioneer Venus (Pioneer 13) data (27). The reanalysis of
Pioneer Venus data showed additional N species (NO
x
), sug-
gesting further chemical disequilibrium in the cloud layers. The
cloud particles themselves also contain many unknowns. The
largest particles, predominant in the lower cloud decks [called
Mode 3 particles (29)], may have a substantial solid component,
implying that they cannot be exclusively made of liquid concen-
trated sulfuric acid (30).
Some additional anomalies that are not directly relevant to
this work, such as the “unknown UV absorber” (31) and the
possible presence of methane (32) or phosphine (33, 34), have
all been suggested as signs of life in the clouds.
How the Rimmer et al. Model Resolves the SO
2
and H
2
O
Abundance Conundrum
Recently, Rimmer et al. (20) proposed a mechanism to explain
the depletion of SO
2
in the atmospheric cloud layers, as well as
the vertical abundance profile of H
2
O in and above the clouds.
If a base is present inside the cloud sulfuric acid droplets, SO
2
will dissolve in the liquid droplets (by reaction with OH
)to
form sulfite. The base (B), therefore traps the SO
2
inside the
cloud droplet as sulfite (HSO
3
),
SO2þH2OþB!SO2þBHþþOH!BHþþHSO3
:
In summary, the equilibrium of the reaction
SO2þH2O$H2SO3
is pulled to the right of the above equation, and S(IV) species
are trapped as sulfite salts through reaction with the base.
Thus, SO
2
is depleted in the cloud layer, compared to the
model with no bases. Eventually, the cloud droplets rain down
to lower atmosphere layers, and the salts dissociate due to
higher temperatures, releasing SO
2
.
Water is consumed in the sulfite-forming reaction, but is
recycled into the lower atmosphere on breakdown of the sul-
fites, which provides a mechanism to explain the water vapor
abundance profile through the clouds. Some water is removed
from the cloud layer, but, because it is replenished by recycling
from below the clouds, the water removal is not absolute, and
so some water remains at the cloud top and in the atmosphere
above the clouds. Thus the Rimmer et al. (20) model predicts
that both SO
2
and H
2
O will be present above the clouds but at
substantially lower abundance than they are below the clouds,
in agreement with observation.
The formation of the sulfite salt within a droplet effectively
neutralizes the acid in the droplet, with the very important out-
come that some of the cloud droplets are much less acidic than
previously thought, with a pH between 1 and 1 (20), instead
of an acidity of approximately 11 (on the Hammett acidity
scale). If correct, the revised pH range of some droplets has a
significance for the habitability of the clouds of Venus that can-
not be overstated. Such a pH range is habitable by terrestrial
extremophiles (35), as compared to the acidity of concentrated
sulfuric acid in which all terrestrial life, and most terrestrial bio-
chemicals, would be destroyed (14).
We argue that the identity of any droplet-neutralizing base is
unknown. Rimmer et al. (20) adopted NaOH as a model base
for their calculations, but noted that iron oxides are a more
physically realistic possibility. In principle, minerals that can
absorb SO
2
could be delivered to the clouds from Venusian vol-
canic eruption, from wind lofting of dust, or from meteoritic
infall. However, it has not been demonstrated that such mecha-
nisms could deliver the very high amount of ∼20 tonnes per
second flux of mineral salts (specifically iron oxides) required
(20).
We are motivated to extend ref. 20’s analysis with the
hypothesis that the neutralizing base that is capturing SO
2
is
locally generated in the clouds. We postulate that NH
3
is the
neutralizing agent for the Venusian cloud droplets, trapping
SO
2
and thus explaining the drop in SO
2
abundance across the
clouds. We are additionally motivated by the tentative in situ
observations of NH
3
in the Venus cloud layers, from both Ven-
era 8 chemical assay (28) and Pioneer Venus probe mass spec-
trometry (27). If present, NH
3
observations cannot yet be
readily explained through any known abiotic planetary pro-
cesses (36). We therefore also explore the possibility that the
NH
3
is biologically produced.
Results
Ammonia as a Neutralizing Agent in the Venusian Cloud Droplets.
We propose NH
3
as the only plausible neutralizing base that
can be generated in situ in the clouds from gas-phase compo-
nents (see SI Appendix, section 1 for further details on potential
neutralizing agents in the cloud layers). The presence of NH
3
,
as with any neutralizing base, leads to chemistry that results in
the SO
2
depletion in the clouds and the observed H
2
O abun-
dance profile, and is consistent with a subset of Mode 3 par-
ticles being nonspherical (i.e., not liquid) and not composed of
pure concentrated sulfuric acid. The presence of NH
3
may also
solve the otherwise unexplained presence of O
2
in the clouds,
especially if the source of NH
3
is biological.
To support our hypothesis that NH
3
could explain the
presence of O
2
within the clouds, we first explore the limited
number of possible chemical reactions that could lead to the
formation of NH
3
in the Venusian atmosphere cloud layer con-
ditions (Table 1).
The most abundant source of nitrogen atoms in the atmo-
sphere of Venus is N
2
gas, so, to make NH
3
,N
2
must be
reduced to NH
3
. The reduction of N
2
to form NH
3
requires a
source of hydrogen atoms, and a source of electrons (reducing
equivalents). Hydrogen atoms are rare in the atmosphere of
Venus. The most abundant gas-phase source of hydrogen atoms
in the atmosphere of Venus is H
2
O, followed by HCl. In order
to generate reducing equivalents, some species must be oxi-
dized. Species available to be oxidized include CO, OCS, SO
2
,
N
2
,H
2
O, and HCl. Phosphorus, if present, will be overwhelm-
ingly present as H
3
PO
4
(34); neither H
3
PO
4
or CO
2
can be fur-
ther oxidized.
2of10 jPNAS Bains et al.
https://doi.org/10.1073/pnas.2110889118 Production of ammonia makes Venusian clouds habitable and explains observed
cloud-level chemical anomalies
Downloaded by guest on December 20, 2021
The most energy- and water-efficient NH
3
-producing reac-
tion (reaction 3 in Table 1) also produces molecular oxygen. We
choose reaction 3, 2N
2(aq)
+10H
2
O
(l)
!4NH
4
+
OH
(aq)
+
3O
2(aq)
, as the basis for our model for two reasons, Firstly, par-
simony leads us to prefer a reaction that uses the smallest
amount of rare materials (H
2
O and energy). Secondly, reaction
3 is the only NH
3
-forming reaction that directly produces O
2
in
the clouds (Table 1), whose detection is one of the anomalies
we wish to explain (discussed below); the other reactions pro-
duce different oxidized species which would not be observed
but which would also produce O
2
on breakdown, and at the
cost of greater energy consumption.
A key question is what NH
3
production rate (by reaction 3) is
needed for maintaining the low SO
2
abundance, as compared to
expected equilibrium values in the atmospheric cloud layers. We
base the SO
2
production rate on the rate at which SO
2
would be
replenished into the clouds by mixing from below, and hence the
rate at which it must be removed from the clouds. The flux is
∼10
11
tonnes per year NH
3
, which is on the order of photosyn-
thetic production of O
2
on Earth (see Materials and Methods).
This flux is calculated assuming that NH
3
is only produced to
sequester SO
2
,andthatonlyNH
3
sequesters SO
2
. If other species
contribute to removing SO
2
, whether hydroxide salts, iron oxides,
or other species, the NH
3
production will be accordingly lower.
AnybyproductofSO
2
sequestration must have a flux of ∼10
11
tonnes per year at the bottom of the clouds, based on the SO
2
depletion within the clouds. A flux of 10
11
tonnes per year is con-
sistent, to within an order of magnitude, with the mass loss at the
bottom of the clouds from rainout of Mode 3 particles from our
calculations (SI Appendix,section2).
All of the NH
3
-producing reactions in the Venusian atmo-
sphere conditions are highly endergonic (Table 1), and so must
be coupled to an energy source if the reactions are to produce
net, “surplus” NH
3
. There are several energy sources that
could, in principle, drive the production of NH
3
. Lightning falls
short by many orders of magnitude of the necessary rate of pro-
duction of NH
3
(SI Appendix, section 7.1 and Table S3), and is
very unlikely to produce both NH
3
and O
2
simultaneously. Sim-
ilarly, UV photochemistry is unlikely to produce NH
3
in more
than trace amounts (SI Appendix, section 7.2), although we
note that the photochemistry of nitrogen species in concen-
trated sulfuric acid has not been explored. Volcanic sources of
NH
3
on Earth are closely associated with organic deposits,
including coal, and also are quantitatively insufficient, even
based on terrestrial rates of volcanic NH
3
production, which
are likely to be much higher than any plausible NH
3
production
on Venus (SI Appendix, section 7.3 and Fig. S4).
The ability to couple chemical energy to drive endergonic
reactions is a universal characteristic of life, and, specifically,
the use of energy to drive the reduction of N
2
to NH
3
in an oxi-
dizing environment is widely found in terrestrial organisms (37,
38). We should therefore consider the possibility that living
organisms in the clouds of Venus are making NH
3
. All of the
NH
3
-producing reactions presented in Table 1 consume water,
which is a rare resource in the clouds of Venus. The energy
expended and water molecules consumed in the process of
making NH
3
must be balanced by an equally powerful benefit
to the organism for this apparently wasteful chemistry. Neutral-
izing the acid to make the droplets habitable is a clear benefit.
We discuss the other, possibly insuperable barriers to the
concept of life in the Venusian clouds below. Here we only
note that the presence of life could explain the observed pres-
ence of NH
3
and O
2
, and later show that it could explain the
observed vertical abundances of H
2
O and SO
2
within and
above the atmospheric cloud layers, and the semisolid nature of
Mode 3 particles. An additional consequence of the NH
3
cloud
droplet chemistry is that the pH of cloud particles with dis-
solved NH
3
must have a pH between 1 and 1, as first shown
by Rimmer et al. (20) for NaOH (Fig. 1).
The Flux of NH
3
Is within the Plausible Biomass Production. The
flux of NH
3
needed to achieve the neutralization effect is not
prohibitive for a realistic biomass within the cloud droplets. We
calculate the biomass required by this model as follows. The
Table 1. Free energy per mole for NH
3
-generating reactions under Venus cloud conditions
Reaction
Free energy of reaction
(kJ/mol)
Free energy required per
mole of surplus NH
3
(kJ/mol)
Water consumed per
surplus NH
3
14N
2(aq)
+11H
2
O
(l)
!
2NH
4
+
OH
(aq)
+
3NH
4
+
NO
3(aq)
1,730 to 2,024 865 to 1,012 6.5
2N
2(aq)
+8H
2
O
(l)
!
2NH
4
+
OH
(aq)
+3H
2
O
2(aq)
1,203 to 1,471 602 to 736 4
32N
2(aq)
+10H
2
O
(l)
fi
4NH
4
+
OH
2
(aq)
+3O
2(aq)
1,000 to 1,306 262 to 343 2.5
44N
2(aq)
+17H
2
O
(l)
+
3HCl
(aq)
!5NH
4
+
OH
(aq)
+3NH
4
+
ClO
4(aq)
1,364 to 1,634 273 to 323 3.4
5N
2(aq)
+6H
2
O
(l)
+3SO
2(aq)
!(NH
4
+
)
2
SO
42-(aq)
+
2H
2
SO
4(aq)
1,193 to 1,313 N/A N/A
Free energies of NH
3
-producing reactions are calculated from refs. 83–85. Ranges are minimum to maximum over a range of pH =3topH=+4 and
temperature from 2 °C to 115 °C. Concentrations of SO
2
and H
2
O are as described in ref. 34. O
2
fractional abundance is assumed to be 10
6
. Table columns
are as follows. First column: reaction number. Second column: possible chemical reaction that produces NH
3
. Third column: free energy of reaction
assuming that NH
3
is accumulated to 2 molar concentration. For the fourth and fifth columns, values were calculated in terms of “surplus NH
3
,”which is
the amount of NH
3
synthesized as NH
4
OH. Fourth column: free energy per mole of “surplus NH
3
”produced. Fifth column: number of water molecules
consumed per “surplus”NH
3
. Reaction 3 (bold type), which produces molecular oxygen as an oxidized byproduct, is the most efficient, in both its use of
energy and its use to water. We note that reaction 4 could produce hypochlorite, chlorite, or chlorate as an oxidized product, but, as perchlorate is
relatively stable and is the weakest oxidizing agent, we have shown this reaction for illustration only. Reaction 5 generates more acid than it consumes,
and so cannot be a source of the base which neutralizes H
2
SO
3
. We also note that reaction 1 and reaction 4 (reactions making nitrate and perchlorate,
respectively) clouds also alternatively explain the presence of O
2
. Nitrate and perchlorate would “rain out”and decompose to N
2
and O
2
or HCl, Cl
2
, and
O
2
, respectively, below the clouds. In situ measurements of NO
x
and ClO
4
abundance in the clouds could rule out these reactions as a potential source of
indirect formation of O
2
.
EARTH, ATMOSPHERIC,
AND PLANETARY SCIENCES
Bains et al.
Production of ammonia makes Venusian clouds habitable and explains observed
cloud-level chemical anomalies
PNAS j3of10
https://doi.org/10.1073/pnas.2110889118
Downloaded by guest on December 20, 2021
production of 10
11
tonnes per year is equivalent to 310
9
gNH
3
per second. Several species of cyanobacteria fix nitrogen at an
average rate of ∼410
7
g per g wet weight biomass per second
(39–41). If life is present in the clouds of Venus, it will not be
terrestrial life; however, if we take these terrestrial organisms
as precedent, 10
11
tonnes per year would be produced by
∼810
15
g wet weight of organism. While this mass might
appear significantly high, it is ∼1/2,000 (0.05%) the biomass of
the Earth (42). This mass translates to ∼1.5% of the mass of
cloud particles in the lower 5 km of the cloud deck (25).
Our model for the production of NH
3
by life is summarized
in Fig. 2.
Toward a Resolution of Venus Atmospheric Anomalies. The incor-
poration of NH
3
in our photochemistry model of the Venusian
atmosphere produces profiles of atmospheric gases that match
the observed abundances of some atmospheric gases better
than existing models of Venus’s atmosphere. Although NH
3
is
an input to our model, no existing Venus photochemical models
include NH
3
(e.g., refs. 22 and 43). In Figs. 3 and 4, we show a
summary of the output of the modeling with NH
3
included,
compared to the same model run without NH
3
and O
2
input,
the latter as reported in ref. 20. The atmospheric photochemis-
try of the clouds was modeled as described in refs. 20 and 34,
and is summarized in Materials and Methods. Specifically, our
model better explains, compared to previous models, 1) the
observed disequilibria in the clouds of Venus; 2) the measured,
but subsequently ignored, abundances of O
2
in the clouds; 3)
the abundance profile of water vapor; 4) the tentative detec-
tions of NH
3
by Venera 8 and Pioneer Venus probes; and 5)
the abundance profile of SO
2
through the cloud layers. To dem-
onstrate how well our model with NH
3
fits the measured data,
we show three model results in Figs. 3 and 4: one model with
NH
3
, one model without NH
3
but with an unphysical arbitrary
depletion rate of SO
2
(a fix common among other models in
order to fit the data), and one model without NH
3
and without
any artificial chemical constraints.
We now turn to each relevant atmosphere anomaly, first
reviewing the data, and then how the presence of NH
3
helps
resolve the anomaly.
O
2
in the clouds is a natural outcome of NH
3
production. Our
model provides an explanation for the presence of O
2
in the
Venus cloud layers. O
2
has been measured via in situ measure-
ments (44, 45). The Pioneer Venus gas chromatography (GC)
reported 43.6 ppm molecular oxygen (O
2
) in the clouds at 51.6
km, 16 ppm below the clouds at 41.7 km, and no detection of
oxygen at 21.6 km (23). The Venera 14 GC detected 18 ppm O
2
average between 35 and 58 km (24). [The Large Neutral Mass
Spectrometer (LNMS) on Pioneer Venus showed a signal of 32
amu, but this was attributed to O
2
ions formed from reaction of
CO
2
in the mass spectrometer (46), and therefore was consid-
ered unreliable. However, we emphasize that this uncertainty
about the source of O
2
is specific to mass spectrometry (47).]
We also note that several ground-based observations attempted
to provide upper limits for the abundance of O
2
above the
clouds (48, 49). The spectroscopic searches for O
2
have been
subjected to varying interpretations (16, 17) and are claimed to
be difficult to reconcile with the in-cloud O
2
abundance
detected by both Pioneer Venus and Venera probes, because
one expects to observe a gradient of O
2
from above to below
the clouds. Such discrepancies can only ultimately be resolved
by new in situ measurements of O
2
in the clouds of Venus.
In the past, the validity of O
2
has been challenged based on
thermodynamics. Initial studies of the atmosphere of Venus in
the 1970s and 1980s assumed the atmosphere was at thermody-
namic equilibrium. One author discounted O
2
as follows (44):
“We therefore conclude, that either we have to accept a strong
disequilibrium state among CO, SO
2
,O
2
and H
2
O in the lower
atmosphere of Venus, or discard at least one of the measure-
ments in order to save the assumption of thermodynamic equi-
librium. The latter course is our preferred one.” Some subse-
quent studies followed this argument (17, 23, 36), although not
all (50), and the author himself modified his opinion in a subse-
quent paper (45). By now, it has been accepted for over two
decades that the atmosphere of Venus is not at thermodynamic
equilibrium (25, 26, 51, 52), although Venus’s atmosphere is
not as far from disequilibrium as Earth’s atmosphere is (51,
52). Recently, the reanalysis of the Pioneer Venus data showed
the atmosphere was farther from equilibrium than previously
thought, due the presence of a range of reduced gases (27).
Still, the cause of the Venus atmosphere thermodynamic dis-
equilibrium is one of the unsolved problems in Venus science
(17).
If the chemistry of NH
3
production is the source of O
2
, then
our model predicts on order 1 ppm O
2
in the cloud level
between about 50 and 60 km; 1 ppm is 20-fold lower than the
measured values (23, 24). However, the value of 1 ppm at lower
altitudes is far greater (15 orders of magnitude) than predicted
by our and other photochemistry models that exclude NH
3
.
While there are no known nonbiological processes that could
produce O
2
locally in the clouds of Venus, we note, for future
work, that other biological processes such as oxygenic photo-
synthesis could also be contributing to the overall O
2
budget in
the clouds.
It has been suggested that O
2
could also be produced by
lightning, which is consistent with O
2
’s presence in and below
but not above the clouds (53). Lightning and coronal discharge
can produce O
2
in a CO
2
+N
2
atmosphere (54). A
thermodynamic-based calculation suggests that the amount of
O
2
possibly produced by lightning is four to five orders of mag-
nitude too low to explain the observations (SI Appendix, section
8.1 and Table S4). However, the efficiency of the production of
O
2
by lightning could be tested experimentally on Earth. It is
possible that all the O
2
detections summarized above were
made as spacecraft fell through high-intensity storm regions
(55), but it seems an unlikely coincidence for two or three sepa-
rate probes to experience storms. In addition, any NH
3
present
in the clouds would be destroyed by the lightning, and only
Fig. 1. Predicted pH profile of cloud particles. The blue shaded region
shows the altitude where clouds are present, from 48 km to 62 km. Note
that the plot extends above and below the cloud tops because there are
plausibly cloud particle populations that extend down to the altitude where
sulfuric acid is sublimated, and up into the mesosphere where sulfuric acid
aerosol evaporation may explain the anomalous SO
2
inversion at 80 km to
100 km. Our model provides no constraints on the composition of the meso-
spheric particles, which may well be composed of pure sulfuric acid.
4of10 jPNAS Bains et al.
https://doi.org/10.1073/pnas.2110889118 Production of ammonia makes Venusian clouds habitable and explains observed
cloud-level chemical anomalies
Downloaded by guest on December 20, 2021
trace amounts would reform (SI Appendix, section 7.1). The
thermal decomposition of H
2
SO
4
to O
2
and SO
2
has been sug-
gested as an industrial process (56), but it is unlikely under
Venus conditions (SI Appendix, section 8.2 and Fig. S5).
At altitudes above the cloud level (∼62 km), no O
2
has been
detected, strongly suggesting a fractional abundance of less than
10
7
(48). Yet, all existing photochemical models predict signifi-
cant molecular oxygen above the clouds (e.g., ref. 22) due to the
instability of CO
2
to photolysis. CO
2
is dissociated into CO and
O, which cannot rapidly recombine because the recombination
reaction is spin forbidden. Some alternative pathway, involving,
for example, OH chemistry, sulfur chemistry, or chlorine chemis-
try, is required to restore CO
2
(see ref. 18), but none of these
pathways are sufficient to draw above-cloud O
2
below 1 ppm (22).
This mismatch between the extremely low observed O
2
levels
above the clouds and the higher predicted levels is a well-known
conundrum of Venus’s cloud layer chemistry. Our model provides
a partial solution by predicting a reduced O
2
level above the
clouds compared to the same model without NH
3
(Figs. 3 and 4).
Model output H
2
O and SO
2
abundance profiles are consistent
with observations. Our photochemistry model with NH
3
production is, together with the model it is based on (20),
consistent with the observed H
2
O and SO
2
abundance profiles
in and above the clouds.
SO
2
and H
2
O have been observed on many occasions by
remote campaigns, orbiters, and in situ probes (reviewed in
refs. 20 and 26). For example, the Visible and Infrared Thermal
Imaging Spectrometer instrument on board Venus Express
observed a mean abundance of H
2
O and SO
2
, below the clouds
at 30 km to 40 km, to be ∼30 ppm and ∼150 ppm, respectively
(57). The observed abundances of H
2
O and SO
2
just below the
clouds are consistent between remote, orbiter, and in situ
observations (20). Recall, that the 5×excess SO
2
over H
2
O
should strip all the water out of the cloud layer, and hence
remove all water above the clouds as well, a solution that is not
consistent with observations. The Rimmer et al. (20) model
uses cloud chemistry (NH
3
or mineral bases) to strip the SO
2
in
the clouds. As a result, water remains in the clouds and above
the clouds, which agrees with the remote, orbiter, and in situ
observations of a few parts per million of H
2
O above the cloud
layers (reviewed in ref. 20).
Within the Venus cloud layers, there is substantial difference
among measurements of water abundance in the clouds as sum-
marized by ref. 20, which may represent varied local conditions.
Fig. 2. Ammonia cycle in the atmosphere of Venus. See SI Appendix, section 10 for details. I: NH
3
is produced locally in the clouds from atmospheric N
2
and H
2
O (Table 1) by metabolically active microorganisms (black dots) inhabiting cloud droplets (white circle). II: The production of NH
3
in the droplet
raises the droplet pH to 1 to 1 (from 11 on the Hammett acidity scale) by trapping the SO
2
and H
2
O in the droplet as ammonium hydrogen sulfite
(NH
4
HSO
3
). The production of sulfite salts in the droplet leads to the formation of a large, semisolid (and hence nonspherical) Mode 3 particle (white
decagon). III: The Mode 3 particle settles out of the clouds where ammonium sulfite disproportionates to ammonium sulfate and ammonium sulfide; the
latter decomposes to H
2
S and NH
3
, which, in turn, undergo photochemical reactions to a variety of products. IV: Disproportionation and gas release break
up the Mode 3 particles into smaller haze particles and microorganism spores (black ovals), some of which return to the cloud layer (V). VI: The ammo-
nium sulfate particles fall farther below the cloud decks, where ammonium sulfate decomposes to SO
3
,NH
3
, and H
2
O. VII: Spores released at this stage
may be unviable (gray ovals), but any surviving could also be eventually transported back to the clouds.
EARTH, ATMOSPHERIC,
AND PLANETARY SCIENCES
Bains et al.
Production of ammonia makes Venusian clouds habitable and explains observed
cloud-level chemical anomalies
PNAS j5of10
https://doi.org/10.1073/pnas.2110889118
Downloaded by guest on December 20, 2021
In models previous to the one described here, water vapor is
removed at the cloud tops by reaction with SO
3
to form sulfuric
acid, which then condenses out to form the cloud droplets.
Since there is more below-cloud SO
2
than H
2
O, all the H
2
O
above the clouds is removed, some models even showing com-
plete depletion of H
2
O (33). Yet, this depleted H
2
O above the
clouds does not match observations which show plenty of water
vapor above the clouds (Figs. 3 and 4). Models previous to ours
solve this problem with physically unrealistic numerical fixes,
either including excess H
2
O below the clouds (21) or fixing the
H
2
O abundance to observed values such that any reactions
involving H
2
O do not consume any H
2
O (22). Most models
avoid the water vapor abundance problem altogether by
restricting the calculations only to a section of the atmosphere,
above the clouds or below the clouds.
Critically important is that our model without NH
3
(33)
must, similarly to other models, impose nonphysical constraints
on SO
2
chemistry in order to make the SO
2
gas abundance pro-
file fit observations, specifically by adding an arbitrary removal
rate for SO
2
in the clouds tuned to fit the data.
We emphasize that our model that includes NH
3
or another
base (20) is the only model known that avoids artificial fixes of
SO
2
and H
2
O. To further emphasize this point, Figs. 3 and 4
include our very poorly fitting model gas abundance profiles
without NH
3
and without the artificial SO
2
removal rate.
Below the clouds, our photochemical model with NH
3
pre-
dicts the same H
2
O abundance as models without NH
3
, includ-
ing previous models (e.g., ref. 43).
NH
3
in the clouds and below the cloud layers is consistent with
tentative observations. NH
3
is a necessary input for our photo-
chemistry model; indeed, the input of NH
3
is the core assumption
of our hypothesis. We therefore discuss the tentative observations
of NH
3
on Venus.
The Venera 8 descent probe reported the presence of NH
3
in the lower atmosphere of Venus. The estimated amounts
from the signal are large and varied from 0.01 to 0.1%. (For
further discussion on the validity of the Venera 8 NH
3
detec-
tion, see SI Appendix, section 6.) A recent reassessment of the
Pioneer Venus LNMS has also provided suggestive evidence
for the presence of NH
3
and its oxidation products in gas phase
in the cloud decks of Venus (27).
The Venera 8 observations were largely discounted at the
time because NH
3
is not likely to be present if Venus’s atmo-
sphere is in thermodynamic equilibrium (36). At least one
author supported the plausibility of the presence of NH
3
in the
cloud layers: Florensky et al. (50), in the late 1970s, argued that
the upper parts of the Venus troposphere do not necessarily
have to be in chemical equilibrium and could contain a number
of minor chemical species, including NH
3
(45).
An additional argument against the plausibility of NH
3
is that
an atmosphere containing sulfuric acid droplets cannot contain a
significant amount of a free base; all of the base, in this case NH
3,
would be sequestered in the droplets as ammonium ions. How-
ever, if the clouds have a pH of >0 and contain significant ammo-
nium salts, then partial pressures of >1 ppm of free ammonia gas
are expected over those droplets in the lower clouds (Materials
and Methods and SI Appendix,section5).
Our model provides a mechanism for the release of NH
3
below the clouds. As the droplets gravitationally settle out of
the atmosphere to higher temperatures, the droplet evaporates,
and NH
3
is released through the thermal decomposition of
ammonium sulfate and ammonium sulfite. NH
3
is subsequently
oxidized to NO
x
and N
2
(Fig. 2). We note that a NO
x
signal has
been identified in the Pioneer Venus LNMS reanalyzed data
(27).
Mode 3 cloud particles. Measurements by the Pioneer Venus
and Venera Probes indicate that the Mode 3 particles might
not be spherical, and that their composition differs from pure
concentrated sulfuric acid. (See SI Appendix,section3for a
brief discussion of the observational support for nonspherical
particles.)
If NH
3
is the main neutralizing agent of the sulfuric acid
cloud droplets, then the Mode 3 cloud particles in the lower
clouds must be supersaturated in ammonium salts, with a small
liquid phase, and therefore are not liquid droplets of concen-
trated sulfuric acid. Thus, the mechanism proposed here pre-
dicts that the Mode 3 particles in the lower cloud are solid or
semisolid, and hence likely to be nonspherical.
Specifically, the Mode 3 (largest) cloud particles in the lower
cloud must be 9.3 molar to 18.1 molar in ammonium salts in
order to provide sufficient downward transport of SO
2
to pro-
duce the observed drop in SO
2
concentration across the clouds
(Materials and Methods and SI Appendix, section 5). Such con-
centrations are not implausible if the Mode 3 particles in the
cloud are actually a semisolid slurry of ammonium salts and sul-
furic acid.
We note that presence of NH
3
creating nonspherical Mode 3
particles is consistent with the Mode 1 and/or Mode 2 particles
being of quite different composition than the Mode 3 particles.
If NH
3
production were the result of biological activity, then
Fig. 3. Venus atmosphere abundance profiles of key molecular species. The
xaxis is the gas fraction by volume, called the mixing ratio. The yaxis is alti-
tude above the surface in kilometers. The lines are gas mixing ratios from
our models: with NH
3
chemistry (solid lines), without NH
3
chemistry (dotted
lines; model in ref. 20), and without NH
3
but with an arbitrary removal rate
for SO
2
in the cloud layers tuned to fit the data (dashed lines; model in refs.
20 and 34). The colored circles show a representative subset of collated
remote and in situ data (error bars not shown) from r efs. 20 (their table 4)
and 33 (their supplementary table S3). Key is that the baseline model predicts
no NH
3
or H
2
S above the 1-ppb level. Models with NH
3
chemistry have very
different H
2
O, SO
2
,O
2
,andH
2
Svaluesatsomealtitudesthanmodelswithout
NH
3
chemistry, and improve the match to observational data. The main take-
away is that the model without NH
3
and without the SO
2
arbitrary removal
rate (dotted line) fits the cloud layer data very poorly, whereas the model
with NH
3
(with no arbitrary constraints; solid line) fits the data much better.
The boundary conditions for surface abundance in the photochemical model
are listed in SI Appendix, Table S6.
6of10 jPNAS Bains et al.
https://doi.org/10.1073/pnas.2110889118 Production of ammonia makes Venusian clouds habitable and explains observed
cloud-level chemical anomalies
Downloaded by guest on December 20, 2021
life could be confined to the larger Mode 3 particles, which
have more volume. If NH
3
was produced by a nonbiological
process, then it would be expected to apply to particles of all
sizes, and not discriminate in favor of Mode 3 particles. How-
ever, the data on particle size and shape is consistent with
Mode 1 and 2 particles being spherical (29).
Our model also explains the presence of the so-called stag-
nant haze layer below the cloud decks (30 km to 47 km alti-
tude) (9). If the large Mode 3 particles are made of mostly solid
ammonium sulfite and ammonium sulfate, then evaporation of
any residual H
2
SO
4
at the cloud base leaves dry solid particles.
The subsequent thermal disproportionation of the remaining
salts generates gas that shatters the particles at the cloud base
(∼100 °Cat∼47 km), and the fragmented particles form the
haze. The haze that settles down and is not mixed back up into
the clouds decomposes at ∼200 °C at the bottom of the stag-
nant haze layer at ∼30 km (Fig. 2). The layered structure and
the altitudes of the boundaries between the layers is therefore a
natural consequence of the ammonia-based cloud chemistry.
See also ref. 7 for a discussion of the composition of the haze
layer.
H
2
S below the clouds. We also note that our model predicts the
presence of H
2
S below the clouds (Figs. 2 and 3). The presence
of H
2
S is consistent with the tentative detection of H
2
S below
the clouds by the Venera 14 GC (24), which is the only in situ
measured abundance value for H
2
S. If NH
3
is present in the
Venus atmosphere, H
2
S is a result of disproportionation of
NH
4
HSO
3
that yields NH
3
,H
2
S, and H
2
O to the atmosphere
below the clouds, and hence is a unique output of our model.
H
2
S was also tentatively identified in the recent reanalysis of
the Pioneer Venus LNMS data (27). H
2
S, however, is a known
volcanic gas on Earth so it is likely produced by volcanoes on
Venus as well.
Discussion
Our model provides a view of the habitability of Venusian
clouds. Concentrated sulfuric acid would make the Venusian
cloud environment both chemically aggressive and extremely
dry (7, 14). Our model removes the issue of extreme acidity for
a subset of cloud particles from consideration.
Our model implies that the Mode 3 cloud particles cannot
be all composed of concentrated H
2
SO
4
. Instead, there has to
be a population of cloud particles that are less acidic and have
a higher pH (between 1 and 1) than concentrated sulfuric
acid. Specifically, our model predicts that the Mode 3 cloud
particles are semisolid ammonium sulfites and sulfates (Fig. 2)
with a pH as high as one (Fig. 1). We emphasize that not all
droplets need to contain semisolid ammonium sulfite and (if
the NH
3
is made by life) ammonia-producing microorganisms.
Relevant to the Mode 3 cloud particles is a recent, indepen-
dent finding that the Mode 3 cloud particle composition is not
primarily sulfuric acid, but instead is consistent with some par-
ticles being ammonium hydrogen sulfate (NH
4
HSO
4
), as also
Fig. 4. Venus atmosphere abundance profiles of three molecular species.
The xaxis is the gas fraction by volume, called the mixing ratio. The yaxis
is altitude above the surface in kilometers. The lines are gas mixing ratios
from our models: with NH
3
chemistry (solid lines), without NH
3
chemistry
(dotted lines; model in ref. 32), and without NH
3
but with an arbitrary
removal rate for SO
2
in the cloud layers tuned to fit the data (dashed lines;
model in refs. 20 and 34). Gray points with error bars are data from obser-
vations tabulated in ref. 20. (Top)O
2
. Our model with NH
3
chemistry
improves upon both the long-standing problem of presence and overabun-
dance of O
2
in the upper atmosphere and the presence of O
2
in the cloud
layers. (Middle)H
2
O. Our model with NH
3
chemistry supports the presence
of water vapor above the cloud layer (>80 km). (Bottom)SO
2
. Our models
with NH
3
chemistry (solid line) and without NH
3
chemistry but with arbi-
trary constraints on SO
2
(dashed line) both provide a fit to observed values
throughout the atmosphere except for the top (>85 km). Key is that the
model without NH
3
and without the SO
2
arbitrary removal rate (dotted
line) fits the cloud layer data very poorly, whereas the model with NH
3
(with no arbitrary constraints; solid line) fits the data much better.
EARTH, ATMOSPHERIC,
AND PLANETARY SCIENCES
Bains et al.
Production of ammonia makes Venusian clouds habitable and explains observed
cloud-level chemical anomalies
PNAS j7of10
https://doi.org/10.1073/pnas.2110889118
Downloaded by guest on December 20, 2021
predicted in our analysis. Mogul et al. (58) base this finding on
a reanalysis of the Pioneer Venus legacy data on the refractive
index of the Venusian cloud droplets, independent of atmo-
spheric chemistry. Also, independently from our work pre-
sented here, Mogul et al. (58) have described the potential for
phototropic synthesis of NH
3
to neutralize sulfuric acid cloud
droplets, leading to the Mode 3 particle possibly containing
NH
4
HSO
4
.
A pH of zero to one is within the range of environments
known from Earth to be habitable and, in fact, to be inhabited.
Life can grow in acid (pH =0) aqueous environments (35), and
microbial growth in solutions as acidic as a pH =0.5 has been
claimed (59). Furthermore, most of the Mode 3 particles have
been detected at altitudes in the temperature range (60 °Cto
80 °C), a range that overlaps with environments known to har-
bor thermophilic acidophiles on Earth (with life that can grow
in temperatures up to 100 °C; e.g., refs. 60–63).
Remarkably, examples of life on Earth secreting NH
3
to neu-
tralize a droplet-sized acidic environment exist. Pathogens such
as Mycobacterium tuberculosis and Candida albicans can neu-
tralize the interior of phagosomes (acid-containing vesicles
inside cells used for digestion of captured organic material) by
secreting ammonia, thus evading destruction (64–66). Some
plant pathogens also secrete ammonia to neutralize local pH in
their target plant cells (67). By contrast, pond-dwelling acido-
philic microorganisms adapt to low pH in other ways, because
it is implausible for them to neutralize an entire river or pond.
Challenges to life in the Venus atmosphere remain. The
extreme aridity of the Venus cloud environment has been well
known for decades (e.g., ref. 68), having been often described
(e.g., refs. 7, 14, and 34), and most recently reviewed in ref. 69,
and remains a significant challenge to life as we know it. Our
model predicts a water vapor abundance mixing ratio of 10
5
in
the lower clouds, that is, a relative humidity of 0.02% (depend-
ing on temperature). This is ∼50-fold lower than the lowest
water activity known to support life on Earth. (We note that
terrestrial life can survive extremely hot and dry environments
as spores or other inactive forms, as summarized in the legend
to Fig. 2 and SI Appendix, section 10, but these are not actively
growing, and to survive an ecosystem requires at least some
cells or organisms to be actively growing.) The range of
in-cloud water vapor abundance mixing ratios reported in the
literature is very large (5 ppm to 0.2%), as summarized by ref.
20, which may represent the presence of more clement local
conditions. All global models may therefore represent an aver-
age of extremely arid “desert” regions and much more humid
“habitable” regions.
The extreme aridity is a reflection of the very low number
density of hydrogen atoms in the Venusian atmosphere. The
scarcity of H atoms argues against the presence of life. Terres-
trial biochemicals are typically ∼50% hydrogen by atom num-
ber (as illustrated by the database of natural products compiled
by ref. 70; SI Appendix, section 9 and Table S5). However,
much of the water in a bacterial cell is derived from reactions
of the metabolites within the cell (71–73). For example, under
active growth of Escherichia coli, up to 70% of the intracellular
water is generated during metabolism and not transported
across the membrane from the outside environment (71). If
there is life on Venus, it is therefore likely to have substantially
different biochemistry from Earth’s, and, if it is based on water
as a solvent, it is likely to have very different strategies for
water accumulation and retention to combat extreme aridity of
the clouds. We note, however, that the lack of hydrogen is not
just a challenge for the habitability of Venus’s clouds but also a
challenge for making detectable amounts of any hydrogen-
saturated gas-phase species, such as NH
3
, by any mechanism,
abiotic or biological.
We note that additional challenges such as nutrient scarcity
or high energy requirements are comparatively less limiting
than aridity; for an in-depth discussion of the challenges to life
in the Venusian clouds, see ref. 7.
An origin for life on Venus is an open question. If life exists
in the Venus clouds, it may have originated on the Venus sur-
face and migrated into the clouds. One model of Venus’s evolu-
tion to its modern state suggests that Venus had clement
surface conditions after formation, only to have entered the
current greenhouse runaway after up to 3.5 billion years (74,
75). This model is dependent on a range of specific conditions
but, if correct, suggests that Venus in the past had similar con-
ditions to those under which life originated on Earth. If life
emerged on the surface, terrestrial precedent suggests that
some organisms would adapt to living some of the time in the
clouds (reviewed in ref. 7). The microbial acid-neutralizing
strategy provides a facile evolutionary path to Venusian cloud
life. As the Venus surface became increasingly hot and uninha-
bitable, cloud dwelling would become a permanent lifestyle.
As the atmospheric chemistry changed to high acidity, the
cloud-dwelling organisms would adapt by neutralizing their
droplet habitats. A plausible evolutionary path is therefore sug-
gested by the unique role of a droplet environment in the acid-
neutralizing strategy, and the proposed history of Venus. We
note, however, that, if life is the source of NH
3
on Venus, it
very likely does not resemble the elemental ratios of life on
Earth and likely has a different biochemistry than life on our
planet, specifically adapted to the unique challenges of the
Venusian cloud environment.
The Venus low D/H ratio (76) and the possible existence of
felsic rocks which form in the presence of water (77–80) imply
the presence of past Venus oceans, yet the debate on whether
or not Venus ever had oceans continues. Recently, Turbet et al.
(81) demonstrated, with a three-dimensional (3D) global cli-
mate model, that Venus may have been too hot early on for
water oceans to form. Their climate model shows that the
steam atmosphere of early Venus never condensed on the plan-
et’s surface to form liquid water oceans. Instead, according to
the model, water vapor condensed on the nightside of the
planet to form clouds that warmed the surface by absorbing
and reemitting the planet’s outgoing infrared radiation (81).
However, Turbet et al. (81) do state that a comprehensive sensi-
tivity study is needed to quantitatively confirm their result, as
cloud and atmospheric circulation feedbacks can vary nonli-
nearly and nonmonotonically with rotation period. The newly
selected VERITAS and EnVision missions, as well as DAVIN-
CI’s instruments, should solidify or rule out the possibility of
the past water-rich era of Venus, by a combination of D/H
measurements and multispectral imaging of the tesserae
regions for mineral compositions.
Summary and Critical Future Measurements
Our hypothesis of locally produced NH
3
in the Venus clouds
explains a number of anomalies in the atmosphere and clouds
of Venus. Our photochemical model of the consequences of
NH
3
production explains the SO
2
depletion in the clouds and
vertical abundance profile of H
2
O, building on the work of ref.
20, explains the presence of O
2
in the clouds, supports the in
situ detection of H
2
S below the clouds, and explains the non-
spherical nature of Mode 3 particles. While the presence of
other mineral bases could contribute, none of them can explain
the parts per million levels of O
2
in the clouds or the tentative
presence of NH
3
. No definitive source for NH
3
has been identi-
fied; in chemical terms, biological production is the most plausi-
ble, but the concept of life in the clouds of Venus remains con-
troversial. Many of the in situ observations should be repeated
8of10 jPNAS Bains et al.
https://doi.org/10.1073/pnas.2110889118 Production of ammonia makes Venusian clouds habitable and explains observed
cloud-level chemical anomalies
Downloaded by guest on December 20, 2021
for confirmation, and more model work is needed to fully
resolve the vertical abundance profiles of relevant gases.
We must be careful not to fall for a conjunction fallacy.
While life may explain the combined anomalies with some
external assumptions, there may yet be a chemical explanation
for each individual anomaly.
An in situ Venus probe can support or refute our proposed
view of Venus as an inhabited planet with the following
measurements.
Gases.
•Establish the existence of NH
3
and O
2
in the cloud layers.
•Measure the amounts of NO
x
to establish which NH
3
-destruc-
tion pathway dominates.
•Determine the specific altitude-dependent abundance profiles
of H
2
O, SO
2
, and H
2
S, ideally with day and night measure-
ments to inform chemistry sources and sinks.
Cloud Particles.
•Confirm the nonspherical, semisolid nature of Mode 3 cloud
particles and identify them as ammonia salts.
•Measure the pH of cloud particles, especially Mode 3 cloud
particles
•Detect organic molecules in cloud particles; if found exclu-
sively in the larger particles, this would be an indicator of life.
Search for Life.
•Analyze a large number of individual cloud particles, espe-
cially Mode 3, for morphological and chemical signs of life.
In the meantime, a public release of original data from the Rus-
sian Venera and Vega missions could enable further support or
refutation of current models and predictions, and would pro-
vide needed context for future mission results.
We have presented an initial analysis of several sources for
the NH
3
on Venus. We have argued that biological production
may be a potential source of both NH
3
and O
2
that we have
identified that meets the quantitative requirements for NH
3
production. Although the biomass required to make NH
3
and
O
2
at the required rate is not unrealistic, at 0.05% of the total
biomass on Earth and ∼1.5% of the total Venusian cloud mass,
life in the clouds of Venus has been considered implausible
because of very high acidity, very low water activity, and scarcity
of hydrogen atoms. By predicting a Mode 3 particle pH of 1
to 1 due to neutralizing NH
3
, our work implies both that Venus
clouds are more habitable than previously thought and, by the
requirement of locally produced NH
3
, that clouds may be
inhabited. We hope our work will encourage further studies
into habitability and astrobiological potential of Venusian
clouds.
Materials and Methods
Photochemical Model. The details of the model are provided in SI Appendix,
section 4. In summary, we employ a 1D Lagrangian photochemistry/diffusion
code that follows a single parcel as it moves from the bottom to the top of the
atmosphere. The temperature, pressure, and actinic UV flux are prescribed at
each altitude in the atmosphere(20).
Calculation of Flux of Ammonia. We calculate the flux of ammonia necessary
to maintain the observed gradient of SO
2
through the clouds following the
method of Rimmer et al. (20). The goal is to explain the removal of nearly all
of 3.510
15
cm
3
of SO
2
(1.510
4
bar at 300-K level of the atmosphere) that
should be present from upward mixing from volcanic sources and recycled
SO
2
. The time taken for SO
2
to mix through the region 45 km to 65 km is calcu-
lated usingLee et al.’s(82)equation7,
Time ≈2H0
:δh
Kzz
¼2:6108s≈8:25 Earth years:
In other words, SO
2
will be replenished in the atmospheric cloud layers in 8.25
Earth years, and this is the timescale thatthe presence of NH
3
needstoremove
SO
2
. The atmospheric scale height H
0
≈6.5 ×10
5
cm is the average scaleheight
in the atmospheric cloud layers, δh=2×10
6
cm is the distance between 45
and 65 km, and K
zz
=10
4
cm
2
s
1
is the eddy diffusion coefficient throughout
the atmospheric cloud layers. The flux (per square centimeter per second) of
SO
2
into the clouds is thereforegiven by
Φ¼AmountDistance
Time ¼3:51015cm3
ð2106cmÞ
2:6108s¼2:7
×1013 cm2s1
:
Recall that there is a one-to-one molar ratio for NH
3
to remove SO
2
. Given the
mass of NH
3
,istheflux rate above is equivalent of 1.1×10
11
tonnes per year of
NH
3
.
Calculation of Concentration of NH
3
in Particles. The concentration of salts in
the cloud droplets can be estimated from the concentration necessary to pro-
vide the flux of NH
3
as calculated above. The necessary flux of NH
3
is depen-
dent on the size of the particles, and hence the particles’rate of fall. For a
given particle size, we can calculate the rate of fall, and hence the volume of
cloud material removed p er second, and, f rom this, the concentr ation of s alts
in that volume needed to provide the flux calculated above. See SI Appendix,
section 2 for the details on the calculation of the concentration of ammonium
salts in the lower cloud particles.
Calculation of Concentration of Gaseous NH
3
over Droplets. The concentra-
tion of gaseous NH
3
over an acid droplet containing dissolved NH
4
+
was calcu-
lated as follows. The fraction of total N species that is present as NH
3
and as
NH
4
+
can be calculated from the acid dissociation constant (pK
a
)ofNH
3
as the
pH. The concentration of NH
3
over solution can be calculated from the solu-
tion concentration and Henry’sconstant(SI Appendix, section 5 and Fig. S3).
Both pK
a
and Henry’s constant are dependent on temperature.
Data Availability. Previously published data were used for this work (20). All
other study data are included in the article and/or SI Appendix.
ACKNOWLEDGMENTS. P.B.R. thanks the Simons Foundation for funding
(Simons Collaboration on the Origins of Life [SCOL] Award 599634). S.S. thanks
the Change Happens Foundation and Breakthrough Initiatives for partial
funding of this work. We are grateful to Vladimir Krasnopolsky for useful dis-
cussionson the presence of O
2
in the atmosphere of Venus.
1. H. Morowitz, C. Sagan, Life in the clouds of Venus? Nature 215, 1259–1260 (1967).
2. C. S. Cockell, Life on Venus. Planet. Space Sci. 47, 1487–1501 (1999).
3. D. Schulze-Makuch, L. N. Irwin, Reassessing the possibility of life on Venus: Proposal
for an astrobiology mission. Astrobiology 2, 197–202 (2002).
4. D. Schulze-Makuch, D. H. Grinspoon, O. Abbas, L. N. Irwin, M. A. Bullock, A sulfur-
based survival strategy for putative phototrophic life in the Venusian atmosphere.
Astrobiology 4,11–18(2004).
5. S. S. Limaye et al., Venus' spectral signatures and the potential for life in the clouds.
Astrobiology 18 (2018), 1181–1198.
6. D. Schulze-Makuch, The case (or not) for life in the Venusian clouds. Life (Basel) 11,
255 (2021).
7. S. Seager et al., The Venusian lower atmosphere haze as a depot for desiccated
microbial life: A proposed life cycle for persistence of the Venusian aerial biosphere.
Astrobiology 21, 1206–1223 (2021).
8. V. A. Krasnopolsky, Photochemistry of the Atmospheres of Mars and Venus (Springer
Science & Business Media, 2013).
9. D. V. Titov, N. I. Ignatiev, K. McGouldrick, V. Wilquet, C. F. Wilson, Clouds and hazes
of Venus. Space Sci. Rev. 214,1–61 (2018).
10. J. M. Jenkins, P. G. Steffes, D. P. Hinson, J. D. Twicken, G. L. Tyler, Radio occultation
studies of the Venus atmosphere with the Magellan Spacecraft. 2. Results from the
October 1991 experiments. Icarus 110,79–94 (1994).
11. T. Imamura et al., Initial performance of the radio occultation experiment in
the Venus orbiter mission Akatsuki Akatsuki. Earth Planets Space 69, 137
(2017).
12. A. T. Young, Are the clouds of Venus sulfuric acid? Icarus 18, 564–582 (1973).
13. J. K. Barstow et al., Models of the global cloud structure on Venus derived from
Venus Express observations. Icarus 217, 542–560 (2012).
14. W. Bains, J. J. Petkowski, Z. Zhan, S. Seager, Evaluating alternatives to water as sol-
vents for life: The example of sulfuric acid. Life (Basel) 11, 400 (2021).
15. N. R. Izenberg et al., The Venus life equation. Astrobiology 21, 1305–1315 (2021).
16. F. P. Mills, M. Allen, A review of selected issues concerning the chemistry in Venus’
middle atmosphere. Planet. Space Sci. 55, 1729–1740 (2007).
EARTH, ATMOSPHERIC,
AND PLANETARY SCIENCES
Bains et al.
Production of ammonia makes Venusian clouds habitable and explains observed
cloud-level chemical anomalies
PNAS j9of10
https://doi.org/10.1073/pnas.2110889118
Downloaded by guest on December 20, 2021
17. V. A. Krasnopolsky, Chemical composition of Venus atmosphere and clouds: Some
unsolved problems. Planet. Space Sci. 54, 1352–1359 (2006).
18. Y. L. Yung, W. B. Demore, Photochemistry of the stratosphere of Venus: Implications
for atmospheric evolution. Icarus 51, 199–247 (1982).
19. C. D. Parkinson et al., Distribution of sulphuric acid aerosols in the clouds and upper
haze of Venus using Venus Express VAST and VeRa temperature profiles. Planet.
Space Sci. 113–114, 205–218 (2015).
20. P. B. Rimmer et al., Hydroxide salts in the clouds of Venus: Their effect on the sulfur
cycle and cloud droplet pH. Planet. Sci. J. 2, 133 (2021).
21. J. R. Winick, A. I. F. Stewart, Photochemistry of SO2 in Venus’upper cloud layers. J.
Geophys. Res. Space Phys. 85, 7849–7860 (1980).
22. C. J. Bierson, X. Zhang,Chemicalcycling in theVenusian atmosphere:A full photochemi-
cal model from the surface to 110 km. J. Geophys. Res. Planets 125, e2019JE006(2020).
23. V. I. Oyama et al., Pioneer Venus gas chromatography of the lower atmosphere of
Venus. J. Geophys. Res. Space Phys. 85, 7891–7902 (1980).
24. L. M. Mukhin et al., VENERA-13 and VENERA-14 gas chromatography analysis of the
Venus atmosphere composition. Sov. Astron. Lett. 8, 216–218 (1982).
25. L. W. Esposito, J. L. Bertaux, V. A. Krasnopolsky, V. I. Moroz, L. V. Zasova, Chemistry of
lower atmosphere and clouds. Venus II, 415–458 (1997).
26. N. M. Johnson, M. R. R. de Oliveira, Venus atmospheric composition in situ data: A
compilation. Earth Space Sci. 6, 1299–1318 (2019).
27. R. Mogul, S. S. Limaye, M. J. Way, J. A. Cordova, Venus’mass spectra show signs of dis-
equilibria in the middle clouds. Geophys. Res. Lett. 48, e2020GL091327 (2021).
28. Y. A. Surkov, B. M. Andrejchikov, O. M. Kalinkina, On the content of ammonia in the
Venus atmosphere based on data obtained from Venera 8 automatic station. Akade-
mia Nauk SSSR Doklady 213, 296–298 (1973),.
29. R. G. Knollenberg, D. M. Hunten, The microphysics of the clouds of Venus: Results of
the Pioneer Venus particle size spectrometer experiment. J. Geophys. Res. Space
Phys. 85, 8039–8058 (1980).
30. R. G. Knollenberg, Clouds and hazes. Nature 296, 18 (1982).
31. S. S. Limaye et al., Venus’spectral signatures and the potential for life in the clouds.
Astrobiology 18, 1181–1198 (2018).
32. T. M. Donahue, R. R. Hodges Jr., Venus methane and water. Geophys. Res. Lett. 20,
591–594 (1993).
33. J. S. Greaves et al., Phosphine gas in the cloud decks of Venus. Nat. Astron. 5, 655–664
(2021).
34. W. Bains et al., Phosphine on Venus cannot be explained by conventional processes.
Astrobiology 21, 1277–1304 (2021).
35. S. Mirete, V. Morgante, J. E. Gonz
alez-Pastor, “Acidophiles: Diversity and mecha-
nisms of adaptation to acidic environments”in Adaption of Microbial Life to Envi-
ronmental Extremes, H. Stan-Lotter, F. Fendrihan, Eds. (Springer, 2017), pp. 227–251.
36. K. A. Goettel, J. S. Lewis, Ammonia in the atmosphere of Venus. J. Atmos. Sci. 31,
828–830 (1974).
37. S. C. Reed, C. C. Cleveland, A. R. Townsend, Functional ecology of free-living nitrogen
fixation: A contemporary perspective. Annu. Rev. Ecol. Evol. Syst. 42, 489–512 (2011).
38. R. H. Burris, G. P. Roberts, Biological nitrogen fixation. Annu.Rev.Nutr.13,317–335 (1993).
39. P. Spr}
ober, H. M. Shafik, M. Pr
esing, A. W. Kov
acs, S. Herodek, Nitrogen uptake and
fixation in the cyanobacterium Cylindrospermopsis raciborskii under different nitro-
gen conditions. Hydrobiologia 506, 169–174 (2003).
40. E. J. Phlips, C. Zeman, P. Hansen, Growth, photosynthesis, nitrogen fixation and car-
bohydrate production by a unicellular cyanobacterium, Synechococcus sp.(Cyano-
phyta). J. Appl. Phycol. 1, 137–145 (1989).
41. D. Vyas, H. D. Kumar, Nitrogen fixation and hydrogen uptake in four cyanobacteria.
Int. J. Hydrogen Energy 20, 163–168 (1995).
42. Y. M. Bar-On, R. Phillips, R. Milo, The biomass distribution on Earth. Proc. Natl. Acad.
Sci. U.S.A. 115, 6506–6511 (2018).
43. V. A. Krasnopolsky, Chemical kinetic model for the lower atmosphere of Venus. Ica-
rus 191,25–37 (2007).
44. U. Von Zahn, S. Kumar, H. Niemann, R. Prinn, “Composition of the Venus atmos-
phere”in Venus, D. M. Hunte, L. Colin, T. M. Donahue, V. M. Moroz, Eds. (University
of Arizona Press, 1983), pp. 299–430.
45. U. Von Zahn, V. I. Moroz, Composition of the Venus atmosphere below 100 km alti-
tude. Adv. Space Res. 5, 173–195 (1985).
46. J. H. Hoffman, R. R. Hodges, T. M. Donahue, M. B. McElroy, Composition of the Venus
lower atmosphere from the Pioneer Venus mass spectrometer. J. Geophys. Res. Space
Phys. 85, 7882–7890 (1980).
47. A. S. Newton, Some observations on the mass spectrum of CO2. J. Chem. Phys. 20,
1330–1331 (1952).
48. J. T. Trauger, J. I. Lunine, Spectroscopy of molecular oxygen in the atmospheres of
Venus and Mars. Icarus 55, 272–281 (1983).
49. F. P. Mills, A spectroscopic search for molecular oxygen in the Venus middle atmo-
sphere. J. Geophys. Res. Planets 104, 30757–30763 (1999).
50. C. P. Florensky, V. P. Volkov, O. V. Nikolaeva, A geochemical model of the Venus tro-
posphere. Icarus 33, 537–553 (1978).
51. D. C. Catling, D. Bergsman, “Using atmospheric composition as a metric for detecting
life on habitable planets“in AGU Fall Meeting Abstracts (American Geophysical
Union, 2009), B11E-06.
52. J. Krissansen-Totton, D. S. Bergsman, D. C. Catling, On detecting biospheres from
chemical thermodynamic disequilibrium in planetary atmospheres. Astrobiology 16,
39–67 (2016).
53. W. L. Chameides, J. C. G. Walker, A. F. Nagy, Possible chemical impact of planetary
lightning in the atmospheres of Venus and Mars. Nature 280,820–822 (1979).
54. D. Nna Mvondo, R. Navarro-Gonz
alez, C. P. McKay, P. Coll, F. Raulin, Production of
nitrogen oxides by lightning and coronae discharges in simulated early Earth, Venus
and Mars environments. Adv. Space Res. 27, 217–223 (2001).
55. M. L. Delitsky, K. H. Baines, Storms on Venus: Lightning-induced chemistry and pre-
dicted products. Planet. Space Sci. 113, 184–192 (2015).
56. General Atomics, “Decomposition of sulfuric acid using solar thermal energy”(Rep.
GA-A-17573, US Department of Energy, 1985).
57. E. Marcq et al., A latitudinal survey of CO, OCS, H
2
O, and SO
2
in the lower atmosphere
of Venus: Spectroscopic studies using VIRTIS-H. J. Geophys. Res. Planets 113, E00B07
(2008).
58. R. Mogul, S. S. Limaye, Y. J. Lee, M. Pasillas, Potential for phototrophy in Venus’
clouds. Astrobiology 21, 1237–1249 (2021).
59. C. Schleper, G. Puhler, H.-P. Klenk, W. Zillig, Picrophilus oshimae and Picrophilus tor-
ridus fam. nov., gen. nov., sp. nov., two species of hyperacidophilic, thermophilic,
heterotrophic, aerobic archaea. Int. J. Syst. Evol. Microbiol. 46, 814–816 (1996).
60. D. Carrizo, L. S
anchez-Garc
ıa, N. Rodriguez, F. G
omez, Lipid biomarker and carbon
stable isotope survey on the Dallol hydrothermal system in Ethiopia. Astrobiology 19,
1474–1489 (2019).
61. D. Rojas-G€
atjens et al., Temperature and elemental sulfur shape microbialcommunities
in twoextremely acidic aquatic volcanic environments. Extremophiles 25,85–99(2021).
62. J. Belilla et al., Hyperdiverse archaea near life limits at the polyextreme geothermal
Dallol area. Nat. Ecol. Evol. 3, 1552–1561 (2019).
63. F. G
omez et al., Ultra-small microorganisms in the polyextreme conditions of the Dal-
lol volcano, Northern Afar, Ethiopia. Sci. Rep. 9, 7907 (2019).
64. H. Song et al., Expression of the ompATb operon accelerates ammonia secretion and
adaptation of Mycobacterium tuberculosis to acidic environments. Mol. Microbiol.
80, 900–918 (2011).
65. A. Gouzy et al.,Mycobacterium tuberculosis exploits asparagine to assimilate nitro-
gen and resist acid stress during infection. PLoS Pathog. 10, e1003928 (2014).
66. S. Vylkova, M. C. Lorenz, Phagosomal neutralization by the fungal pathogen Candida
albicans induces macrophage pyroptosis. Infect. Immun. 85, e00832-16 (2017).
67. N. Alkan, R. Fluhr, A. Sherman, D. Prusky, Role of ammonia secretion and pH modula-
tion on pathogenicity of Colletotrichum coccodes on tomato fruit. Mol. Plant
Microbe Interact. 21, 1058–1066 (2008).
68. T. M. Donahue, R. R. Hodges Jr., Past and present water budget of Venus. J. Geophys.
Res. Planets 97, 6083–6091 (1992).
69. J. E. Hallsworth et al., Water activity in Venus’s uninhabitable clouds and other plane-
tary atmospheres. Nat. Astron. 5, 665–675 (2021).
70. J. J. Petkowski, W. Bains, S. Seager, An apparent binary choice in biochemistry:
Mutual reactivity implies life chooses thiols or nitrogen-sulfur bonds, but not both.
Astrobiology 19, 579–613 (2019).
71. H. W. Kreuzer-Martin, J. R. Ehleringer, E. L. Hegg, Oxygen isotopes indicate most
intracellular water in log-phase Escherichia coli is derived from metabolism. Proc.
Natl. Acad. Sci. U.S.A. 102, 17337–17341 (2005).
72. H. W. Kreuzer-Martin, M. J. Lott, J. R. Ehleringer, E. L. Hegg, Metabolic processes
account for the majority of the intracellular water in log-phase Escherichia coli cells
as revealed by hydrogen isotopes. Biochemistry 45, 13622–13630 (2006).
73. H. Li et al., Probing the metabolic water contribution to intracellular water using oxy-
gen isotope ratios of PO4. Proc. Natl. Acad. Sci. U.S.A. 113, 5862–5867 (2016).
74. M. J. Way et al., Was Venus the first habitable world of our solar system? Geophys.
Res. Lett. 43, 8376–8383 (2016).
75. M. J. Way, A. D. Del Genio, Venusian habitable climate scenarios: Modeling Venus
through time and applications to slowly rotating Venus-like exoplanets. J. Geophys.
Res. Planets 125, e2019JE006276 (2020).
76. T. M. Donahue, J. H. Hoffman, R. R. Hodges, A. J. Watson, Venus was wet: A measure-
ment of the ratio of deuterium to hydrogen. Science 216, 630–633 (1982).
77. N. Mueller et al., Venus surface thermal emission at 1 μm in VIRTIS imaging observa-
tions: Evidence for variation of crust and mantle differentiation conditions. J. Geo-
phys. Res. Planets 113, E00B17 (2008).
78. A. T. Basilevsky et al., Geologic interpretation of the near-infrared images of the sur-
face taken by the Venus Monitoring Camera, Venus Express. Icarus 217, 434–450
(2012).
79. M. S. Gilmore, L. S. Glaze, “The oldest rocks on Venus: The importance of tessera ter-
rain for Venus exploration“in AGU Fall Meeting Abstracts (American Geophysical
Union, 2013), P34A-01.
80. G. L. Hashimoto et al., Felsic highland crust on Venus suggested by Galileo near-
infrared mapping spectrometer data. J. Geophys. Res. Planets 113, E00B24 (2008).
81. M. Turbet et al., Day-night cloud asymmetry prevents early oceans on Venus but not
on Earth. Nature 598, 276–280 (2021).
82. G. Lee, C. Helling, I. Dobbs-Dixon, D. Juncher, Modelling the local and global cloud
formation on HD 189733b. Astron. Astrophys. 580, A12 (2015).
83. J. P. Amend, E. L. Shock, Energetics of overall metabolic reactions of thermophilic
and hyperthermophilic Archaea and bacteria. FEMS Microbiol. Rev. 25, 175–243
(2001).
84. A. A. Gilliland, W. H. Johnson, Heats of formation of lithium perchlorate, ammonium
perchlorate, and sodium perchlorate. J. Res. Natl. Bur. Stand., A Phys. Chem. 65A,
67–70 (1961).
85. R. Meyer, J. K €
ohler, A. Homburg, Explosives (John Wiley, 2016).
10 of 10 jPNAS Bains et al.
https://doi.org/10.1073/pnas.2110889118 Production of ammonia makes Venusian clouds habitable and explains observed
cloud-level chemical anomalies
Downloaded by guest on December 20, 2021
